Arbitrarily stopping the oscillator of a microcontroller may lead to a
host of odd behavior.

Some designs may accept it well, but others may randomly stop a PWM
signal in the 'on' state and very bad things may happen.
The firmware may become confused by the "time lapse" and behave erratically=
..

Also, during the oscillator re-start-up it may produce spurious pulses
that may lead to the CPU hanging. MCUs usually have a start-up delay
timer to wait for the clock stabilization before it starts executing
instructions.

Some MCUs have clock failure detection and may switch to an internal
clock and do unexpected things while its main clock is stopped.

And so on...


Cheers,

Isaac





Em 02/10/2015 09:32, Art escreveu:
> Hi Marino,
>
> What type of mouse is it???
>
> My mouse does an initialization when it is first powered up, drawing 30=20
> mA for about 10 seconds. I'm not sure a fet across the power switch is a=
=20
> good idea, mostly because mosfets don't like slow rising input=20
> transitions and an RC time constant will make the gate voltage=20
> transition horribly slow.
>
> My idea to defeat the oscillator after the wake up timer expires leaves=20
> the mouse in a mode that consumes near zero power AND that allows it to=20
> continue scanning almost instantaneously when the wake up timer=20
> indicates mouse movement.
>
> I have some micropower comparators on order and a mess of sensors.. Time=
=20
> will tell which (if any) method works.
>
> Any other suggestions?? Seriously, I appreciate any input!
>
> Thanks all suggestions.
>
> AG
>
>
>
> On 10/01/2015 08:30 PM, Isaac Marino Bavaresco wrote:
>> I just turned my mouse off and on again, and immediately moved it. I
>> could not notice any turn-on delay.
>>
>> I think that a good design could be using an N-Channel MOSFET on the
>> negative battery wire (or a P-Channel on the positive wire) and a piezo
>> transducer to its gate, with a very large resistor from gate to source.
>> The resistor and the gate capacitance would create an R/C timer that
>> keeps the mouse on for a while and then turn it off automatically. A
>> movement would make the piezo transducer to produce a voltage spike that
>> charges the gate capacitance again.
>>
>> This circuit would "harvest" energy from users movement to keep the
>> mouse on and would not use any quiescent energy from the battery.
>>
>> Perhaps an external capacitor and a Zener diode may be necessary.
>>
>>
>> Cheers,
>>
>> Isaac
>>
>>
>> Em 01/10/2015 17:39, James Cameron escreveu:
>>> On Thu, Oct 01, 2015 at 08:02:45AM -0400, Art wrote:
>>>> So, my question is WHAT METHOD DO I USE TO STOP THE MOUSE FROM DRAWING
>>>> CURRENT FROM THE BATTERY WHEN THE WAKE UP SENSOR TIMES OUT??
>>> No idea, it would depend on the design of the existing circuit, and
>>> the microcontroller code already in place.
>>>
>>> You might also consider replacing the battery with something of
>>> greater energy density.
>>>
>>



--=20
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
.